Spectral patterns of main classes

Due to the limited availability of meta data, we apply basic image processing strategies e.g. contrast stretching, Lillesand and Kiefer, 1999 to the raw intensity values. For each channel, the 99.99th percentile is computed and used as intensity cutoff threshold. This way, outliers from specular reflections are eliminated. Visual inspection of these points revealed that such outliers occur on moving objects like cars and other man-made objects like building facades. Finally, a histogram stretch to twice the standard deviation is computed for each channel, resulting in value ranges from 0 to 255. Figure 1. Merged point cloud coloured as false-colour composite red=C2, green=C3 and blue=C1 The scaled intensity values are used to create false color composites for each point. The combination red=C2, green=C3 and blue=C1 is used to simulate a CIR color infrared similar visual appearance, and the combination red=C1, green=C2 and blue=C3 for a true color similar appearance. The channels C2 and C3 are used to calculate a pseudo NDVI Eq. 1: pseudo NDVI = 1 where pseudo NDVI = pseudo normalized difference vegetation index C2 = near infrared channel 1,064 nm C3 = green channel 532 nm Finally, the point cloud is classified into the main classes ‘unsealed ground’, ‘sealed ground’, ‘buildings’, ‘mid vegetation’, ‘high vegetation’, ‘water surface’ and ‘water body bottom’ and for some small training areas into the subclasses ‘green grass’, ‘dried up grass’, ’sand bare soil’, ’wetlands’, ‘darker asphalt’ and ‘lighter asphalt’. The classification is done by a two stage approach consisting of automatic and semi- automatic classification. First the point cloud is automatically classified into ‘ground’, ‘buildings’, and ‘mid vegetation’ and ‘high vegetation’ classes by a “geometrical classification” using information about x,y,z coordinates and neighborhood exclusively. Ground points are classified using a hybrid approach of progressive TIN densification Axelsson, 2000 and RANSAC-based point cloud segmentation. The latter is also used together with other point cloud features such as eigenvalue based omnivariance, Mallet et al., 2011 to differentiate the remaining non-ground points into buildings and vegetation. For the accuracy assessment of the automatic classification, two test areas, each 600 x 600 m, were selected and manually revised. In a second semi-automatic step, the ground class is further subclassified by introducing three additional classes: i ‘sealed ground’ e.g. roads, ii ‘water surface’, and iii ‘water body bottom’. For further analysis of small training samples the heterogeneous classes of ‘sealed’ and ‘unsealed ground’ were semi-automatically classified into ‘green grass’, ‘dried up grass’, ’sand bare soil’, ’wetlands’, ‘darker asphalt’ and ‘lighter asphalt’. For each channel, histograms showing the intensity value distribution for the seven classes are calculated and analyzed in order to evaluate the potential of extending the geometrical classification approaches currently available for airborne LiDAR mapping. Because of their heterogeneity, the ground classes unsealed and sealed are examined in greater detail. Therefore, training areas composed of different materials are manually selected and again histograms are calculated for each sub-class. The channel peak values of each sub-class are then used for a supervised classification by pattern matching based on the Mahalanobis distance.

4. RESULTS

For the merged channel datasets we computed false color composite point clouds as shown in Fig. 1. Based on the manual classification of the datasets, we looked at the 8 bit scaled channel histograms and the pseudo NDVI for the main classes. As the majority of the data showed only a spread of 201, 167 and 133 integer values for the channels C1, C2 and C3, the 8 bit scaled histograms contain data gaps Fig. 2,3 and 5,6.

4.1 Spectral patterns of main classes

Looking at the two subareas, a set of separated peaks is visible in the histograms of the three channels. This indicates heterogeneity of spectral patterns within one object class Fig. 2 and 3. The subareas are decomposed into the classes ‘unsealed ground’, ‘sealed ground’, ‘buildings’, ‘mid vegetation’, ‘high vegetation’, ‘water surface’ and ‘water body bottom’. Except for the class ‘unsealed ground’, these classes show single peak distributions. The sealed ground contains a peak at low intensities ~50 for C2 and a peak at medium intensities for C1 and C3 ~100, 120. The peaks in the ‘buildings’ class show low intensities for all channels ~50, 25, 50. For both classes this is in agreement with corresponding low pseudo NDVI values. The distributions of ‘mid vegetation’ and ‘high vegetation’ exhibit also peaks at low intensities ~100, 50, 40 ~50, 25, 40 with slightly higher intensities in the mid vegetation. Except for the ‘buildings’ class all classes show a larger spread of intensity values. Besides the main peak, a second peak of drop-outs intensity is zero is observable in all channels. Looking at the ‘water body bottom’ class a unique pattern with two drop-out channels C1 and C2 and the third channel C3 with wide spread intensities can be observed. This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper. Editors: S. Oude Elberink, A. Velizhev, R. Lindenbergh, S. Kaasalainen, and F. Pirotti doi:10.5194isprsannals-II-3-W5-113-2015 115 Figure 2. Class specific 8 bit scaled channel histograms for 1,550 nm and 1,064 nm wavelength laser signals countsbin The ‘unsealed ground’ includes a variety of sub-classes representing ground vegetation of varying health and density. This is most obvious in C2 showing multiple peaks. C2 shows the best contrast to the ‘sealed ground’ class, while C1 and C3 show distributions very similar to the ‘sealed ground’ distributions. The pseudo NDVI also shows a good discrimination between ‘sealed ground’ low NDVI and ‘unsealed ground’ high NDVI. The suitability of the different channels for the separation of ‘sealed ground’ and ‘unsealed ground’ is shown in Fig. 4.

4.2 Spectral patterns of ground sub-class